[1]谢七月,钟树红,戴 杨,等.基于全天空图像关键云图特征的小时内光伏功率预测[J].机械与电子,2026,44(04):33-39.
　XIE Qiyue,ZHONG Shuhong,DAI Yang,et al.Intra-hour Photovoltaic Power Prediction Based on Key Cloud Features from All-sky Images[J].Machinery & Electronics,2026,44(04):33-39.
点击复制

基于全天空图像关键云图特征的小时内光伏功率预测()
分享到：

《机械与电子》[ISSN:1001-2257/CN:52-1052/TH]

卷:
44
期数:
2026年04期
页码:
33-39
栏目:
智能检测
出版日期:
2026-04-27

文章信息/Info

Title:
Intra-hour Photovoltaic Power Prediction Based on Key Cloud Features from All-sky Images
文章编号:
1001-2257(2026)04-0033-07
作者:
谢七月1; 2; 钟树红1; 2; 戴 杨1; 2; 周育才1; 付 强1; 2; 王晓丽3
(1.长沙理工大学人工智能学院,湖南 长沙 410114;
2.电网防灾减灾国家重点实验室,湖南 长沙 410114;
3.中南大学自动化学院,湖南 长沙 410083)
Author(s):
XIE Qiyue1; 2; ZHONG Shuhong1; 2; DAI Yang1; 2; ZHOU Yucai1; FU Qiang1; 2; WANG Xiaoli
(1.School of Artificial Intelligence,Changsha University of Science and Technology,Changsha 410114,China;
2.State Key Laboratory of Power Disaster Prevention and Mitigation,Changsha 410114,China;
3.School of Automation,Central South University,Changsha 410083,China)
关键词:
光伏功率预测; 全天空图像; 智能优化算法; 深度学习; 图像特征提取
Keywords:
photovoltaic power forecasting; all sky images; intelligent optimization algorithms; deeplearning; image feature extraction
分类号:
TM615
文献标志码:
A
摘要:
针对天空云层变化导致光伏功率预测精度下降,以及现有基于全天空图像方法在关键云区特
征提取方面存在不足的问题,提出一种基于全天空图像关键云区特征提取的改进光伏功率预测方法。首
先,提取全天空图像关键云区特征与历史光伏功率特征以捕获扰动与时序信息;随后,采用双通道长短期记
忆网络(LSTM)分别编码过去与未来特征,并通过注意力机制实现高效融合;最后,引入逐步训练策略,并利
用常春藤算法(IVYA)优化模型参数。实测数据表明,所提方法优于空间卷积神经网络(SCNN)、Conv-
NeXt、双向长短期记忆网络(Bi LSTM)及Transformer等主流方法,其10 min预测性能取得EMAE 为
0.515 kW、ERMSE 为1.489 kW、R2 为0.967 8的结果。
Abstract:
To address the degradation in photovoltaic (PV) power forecasting accuracy caused by dynamic
cloud evolution and the insufficient extraction of key cloud region features in existing all sky image–
based methods,an improved PV power prediction approach based on key cloud region feature extraction
from all sky images is proposed.Firstly,the features of key cloud regions from all sky images
and historical PV power are extracted to capture perturbation and temporal information.Subsequently,a
dual channel long short term memory (LSTM) network is employed to encode past and future features
separately,which are then efficiently fused via an attention mechanism.Finally,a stepwise training strategy
is introduced,and the Ivy Algorithm (IVYA) is used to optimize model parameters.Validation using one
year of measured data demonstrates that the proposed method outperforms mainstream models including
SCNN,ConvNeXt,Bi LSTM,and Transformer,achieving a 10 minute forecasting performance with a
EMAE of 0.515 kW,a ERMSE of 1.489 kW,and R2 of 0.967 8.

参考文献/References:

[1] 刘海鹏,何艳苹,金怀平,等.考虑数据分解和Gish BiTCN MHSA的短期光伏功率预测[J] .太阳能学
报,2025,46(8):430 438.
[2] YU S W,HU X,LI L X,et al.Does the development of renewable energy promote carbon reduction? Evidence
from Chinese provinces[J].Journal of environmental management,2020,268:110634.
[3] GANDHI O,KUMAR D S,RODRÍGUEZ GALLEGOS C D,et al.Review of power system impacts at high PV
penetration Part I:factors limiting PV penetration[J].Solar energy,2020,210:181 201.
[4] 商立群,李洪波,侯亚东,等.基于VMD ISSA KELM的短期光伏发电功率预测[J].电力系统保护与控制,
2022,50(21):138 148.
[5] 韩博,李长青,刘卫亮,等.基于二次分解和BiGRU 的超短期光伏发电功率预测[J].动力工程学报,2025,45
(1):62 69.
[6] DI LEO P,CIOCIA A,MALGAROLI G,et al.Advancements and challenges in photovoltaic power forecasting:
a comprehensive review[J].Energies,2025,18(8):2108.
[7] 余光正,陆柳,汤波,等.基于云图特征提取的改进混合神经网络超短期光伏功率预测方法[J].中国电机工程
学报,2021,41(20):6989 7003.
[8] ZANG H X,CHEN D H,LIU J X,et al.Improving ultra short term photovoltaic power forecasting using
a novel sky image based framework considering spatial temporal feature interaction[J].Energy,2024,
293:130538.
[9] 吐松江·卡日,吴现,马小晶,等.基于地基云图数据多维特征融合的光伏功率预测算法[J].电力系统保护与
控制,2025,53(11):84 94.
[10] ARRAIS J M,CERENTINI A,MARTINS B J,et al.Systematic review on ground based cloud tracking
methods for photovoltaics nowcasting[J].American journal of climate change,2024,13(3):452 476.
[11] 姚钦才,向文国,陈时熠,等.基于ICEEMDAN KPCA ICPA LSTM 的光伏发电功率预测[J].动力工
程学报,2025,45(3):374 382.
[12] FENG C,ZHANG J,ZHANG W Q,et al.Convolutional neural networks for intra hour solar forecasting
based on sky image sequences[J].Applied energy,2022,310:118438.
[13] XIE Q Y,MA L,LIU Y,et al.An improved SSA BiLSTM based short term irradiance prediction
model via sky images feature extraction[J].Renewable energy,2023,219:119507.
[14] ZHAO X,FAN H,GUO S,et al.Sky image informed hybrid model for ultra short term prediction of
photovoltaic power output[J].IET Conference proceedings,2024,2024(33):375 379.
[15] WANG Z G,WANG L,HUANG C,et al.A hybrid ensemble learning model for short term solar irradiance
forecasting using historical observations and sky images[J].IEEE Transactions on industry applications,
2022,59(2):2041 2049.
[16] GHASEMI M,ZARE M,TROJOVSK?Y P,et al.Optimization based on the smart behavior of plants with
its engineering applications:Ivy algorithm[J].Knowledge based systems,2024,295:111850.
[17] NIE Y H,LI X T,SCOTT A,et al.SKIPP’D:a sky images and photovoltaic power generation dataset for
short term solar forecasting[J].Solar energy,2023,255:171 179.
[18] LIU Z,MAO H Z,WU C Y,et al.A convnet for the 2020s[C]∥2021 IEEE/CVF Conference on Computer
Vision and Pattern Recognition(CVPR).New York:IEEE,2022:11976 1198.

备注/Memo

备注/Memo:
收稿日期:2025-12-11
基金项目:国家自然科学基金资助项目(62373067)
作者简介:谢七月 (1980-),男,江西于都人,博士,教授,研究方向为发电过程建模与优化控制、机器学习与人工智能。
更新日期/Last Update: 2026-05-11